Description

A flaw has been found in GH05TCREW PentestAgent up to cf882dabea3ed91cef016cdd115e5426315665a2. This issue affects the function LocalRuntime.execute_command of the file runtime/runtime.py of the component LocalRuntime. Executing a manipulation can lead to os command injection. The attack may be performed from remote. The exploit has been published and may be used. The pull request to fix this issue awaits acceptance.

Executive Summary

Technical Details

Mitigations

References

Additional Metadata

{
  "score": "7.3",
  "pubDate": "2026-09-14T03:16:37.263Z",
  "pubdate": "2026-09-14T03:16:37.263Z",
  "executiveSummary": "A critical OS command injection vulnerability exists within the GH05TCREW PentestAgent project, specifically affecting the LocalRuntime component.\nThe vulnerability allows unauthenticated remote attackers to execute arbitrary system commands on the host operating system with the privileges of the application process.\nThis flaw stems from improper input sanitization within the LocalRuntime.execute_command function, which improperly processes untrusted input before passing it to the system shell.\nThe risk implication is severe, as successful exploitation results in full remote code execution (RCE), potentially leading to complete system compromise, data exfiltration, or lateral movement within the network.\nThe vulnerability is currently unpatched in the affected codebase up to commit cf882dabea3ed91cef016cdd115e5426315665a2, and public exploit code exists, increasing the risk of active exploitation.\nGiven the nature of the application as a pentesting tool, the presence of an RCE vulnerability significantly undermines the security integrity of the infrastructure where it is deployed.",
  "technicalDetails": "The vulnerability is located in the LocalRuntime.execute_command function within the file runtime/runtime.py. The root cause is the improper handling of user-supplied input when invoking system-level commands.\nIn the affected component, the application accepts input that is subsequently concatenated or passed directly into a shell execution environment (e.g., os.system or similar subprocess call without secure argument separation).\nBecause the function fails to sanitize or validate input against a whitelist of expected command structures, an attacker can supply malicious characters (such as semicolons, pipes, or command substitution operators) to break out of the intended command context.\nThe attack flow proceeds as follows: 1) The attacker identifies a network-exposed entry point that interacts with the LocalRuntime component. 2) The attacker crafts a malicious payload containing shell meta-characters designed to terminate the intended command and execute an arbitrary secondary command. 3) The attacker submits this payload to the affected PentestAgent interface. 4) The LocalRuntime.execute_command function processes the malicious input and passes the unsanitized string to the underlying system interpreter. 5) The operating system executes the attacker's injected command with the process context of the PentestAgent.\nThe vulnerability is present in versions up to commit cf882dabea3ed91cef016cdd115e5426315665a2. Exploitation does not require prior authentication, and the capability to trigger the function remotely allows for weaponized payloads over network protocols.\nUpon successful injection, the attacker can achieve post-exploitation objectives such as spawning reverse shells, installing persistent backdoors, or accessing sensitive environment variables and configuration files. Since the vulnerability resides in a core component responsible for command execution, the impact is comprehensive, effectively granting the attacker the same operational permissions as the PentestAgent process.\nThe existence of public exploit code confirms that the attack vector is well-understood and trivial to execute for unauthorized parties."
}